Metabolic phenotyping and systems biology approaches to understanding neurological disorders

Abstract

The development of high-throughput metabolic profiling and the study of the metabolome are particularly important in brain research where small molecules or metabolites play fundamental signalling roles: neurotransmitters, signalling lipids, osmolytes and even ions. Metabolic profiling has shown that metabolic perturbations in the brain go beyond alterations of neurotransmission and that variations in brain metabolic homeostasis are associated with neurological disorders. In this report, we will focus on recent developments in the field of metabolic phenotyping that have contributed to unravelling the pathophysiology of neurological diseases. Also, we will highlight the necessity of implementing systems biology approaches to integrate metabolic data and tackle the structural and functional complexity of the brain in normal and pathological conditions.

Figure 1.. Typical metabonomics workflow

Standardized samples collection (A) is followed by NMR or mass spectrometry data acquisition (B). Multivariate statistical analysis of spectral data enables classification and/or prediction of case and control samples based on statistical scores. Model coefficients or loadings (black triangles) are derived to highlight the single or group of peaks assigned to metabolites (here, to simplify representation, only 5 metabolites are represented: m1 to m5) whose variations maximize the discrimination between case and control samples (C). Pattern recognition statistical methods define disease-specific metabolic signatures (D), which correspond to the pattern of significantly affected metabolites (purple dots) representing candidate biomarkers of the disease.

Figure 2.. Principles of integrated metabolome and interactome mapping (iMIM): a biomolecular GPS to navigate the underlying signalling networks and understand metabolic signatures

Disease-causing candidate genes (A) are analysed in relation to patterns of significant metabolites (B) via protein-protein and metabolite-protein interactions (C). In iMIM, network metrics are used to define shortest paths and pivotal proteins between metabolites and disease-causing genes (D).